Embed Size (px)
Citation preview
- 1 -
The effectiveness of conventional trickling filter treatment plants at
reducing concentrations of copper in wastewaters.
Dominik Ziolkoa, David Hala
a, John N. Lester
b and Mark D. Scrimshaw
a,*
a Institute for the Environment, Brunel University, Uxbridge, Middlesex, UB83 PH, UK
b School of Applied Science, Centre for Water Science, Cranfield University, Bedfordshire,
MK43 0AL, UK
* Corresponding author. Tel.: +44 (0)1895 267 299. E-mail: [email protected].
Abstract
Eight different sewage treatment works were sampled in the North West of England.
The effectiveness of the conventional treatment processes (primary sedimentation and
biological trickling filters) as well as various tertiary treatment units in terms of both
total and dissolved copper removal was evaluated. The removal of total copper across
primary sedimentation averaged 53% and were relatively consistent at all sites,
however, at three sites the removal of dissolved copper also occurred at this stage of
treatment. Removal of total copper by the biological trickling filters averaged 49%,
however, substantial dissolution of copper occurred at two sites, which highlighted the
unpredictability of this treatment process in the removal of dissolved copper. Copper
removal during tertiary treatment varied considerable even for the same treatment
processes installed at different sites, primarily due to the variability of insoluble
copper removal, with little effect on copper in the dissolved form being observed. The
proportion of dissolved copper increased significantly during treatment, from an
average of 22% in crude sewages to 55% in the final effluents. There may be the
potential to optimise existing, conventional treatment processes (primary or biological
treatment) to enhance dissolved copper removal, possibly reducing the requirement
for installing any tertiary processes specifically for the removal of copper.
Keywords: copper; sewage; removal efficiency; mass balance; primary treatment;
trickling filters
2
1. Introduction
Copper is one of the metals that are of particular concern in relation to water
quality, because although an essential element, elevated concentrations in the aquatic
environment may result in toxicological effects (de Oliveira-Filho et al., 2004; de
Schamphelaere et al., 2004). Copper has been classified under the Water Framework
Directive as a potential main pollutant (Annex VIII) for which an environmental
quality standard (EQS), if required, should be established at the national level
(European Commission, 2000). At present, in England and Wales for water bodies
achieving good ecological status, the freshwater EQS for copper ranges from 1-28
μg/l (dissolved copper annual average) depending on water hardness, falling to 5 μg/l
in coastal and estuarine areas (UKTAG, 2008). Discharge consents for sewage
treatment works (STW) however, are set based on the total copper concentrations
(particulate plus dissolved copper). Removal of both particulate and dissolved copper
fractions during sewage treatment is therefore relevant to achieve compliance with
increasingly stringent consents and EQS and achieving this with existing unit
treatment processes may be challenging (Ziolko et al., in press).
In developed countries concentrations of copper in sewage have generally
decreased in comparison to the late 1970s (Ziolko et al., in press). However,
contemporary concentrations remain significant when compared with EQS and
consents, especially for receiving waters that are soft in nature and offer limited
natural dilution for final effluents. Thus STW, although not designed to do so, are
expected to reduce the load of copper discharged to the aquatic environment.
However, most contemporary data available on copper concentrations in sewage
(Sorme and Lingerkvist, 2002; Karvelas et al., 2003; Buzier et al., 2006; Rule et al.,
2006; Oliveira et al., 2007) focus on total metal, leaving a paucity of information on
3
the dissolved fraction, despite phase distribution being recognised as a significant
factor in determining the mechanism and extent of copper removal in STW (Kempton
et al., 1983; Rudd et al., 1983a; Lawson et al., 1984a; Kempton et al., 1987a, b).
There are effectively two treatment stages in conventional sewage treatment,
primary sedimentation and biological treatment, where copper (Lester et al., 1979),
and other elements (Stoveland et al., 1979a) may be removed. During primary
treatment, removal is physical, and depends upon the presence of metal in an
insoluble and settleable form (Kempton et al., 1987a, b). Two factors that have the
biggest impact on metal removal at this stage are therefore copper partitioning
between the particulate and dissolved phases in crude sewage and the efficiency of
suspended solids removal (Lawson et al., 1984b, c, d; Kempton et al., 1987a).
Typically no removal of dissolved copper is reported at this stage of treatment and as
a result the proportion of dissolved to particulate copper in the settled sewage
increases (Olivier and Cosgrove, 1974; Stoveland and Lester, 1980; Santarsiero et al.,
1998; Karvelas et al., 2003).
Secondary biological treatment is predominantly based on one of two
processes, a biological reactor consisting of either suspended growth (activated
sludge) or an attached growth process (trickling filters) each followed by a phase
separator (secondary sedimentation). In the activated sludge process the majority of
the biomass removed in the secondary sedimentation tank is recycled to the aerator as
returned activated sludge (RAS) a defining characteristic of the process (Metcalf and
Eddy, 2004). Biological treatment processes are theoretically capable of removing
both dissolved copper, mostly through sorption to the extracellular polymeric
substances (EPS) produced by bacteria, as well as copper associated with solids,
through settling with sludge flocs (Brown and Lester, 1979; Rudd et al., 1983b; Liu et
4
al., 2001). However, due to the strong affinity of copper for dissolved organic ligands
and/or possibility of release of EPS from the bacterial cell capsules, an increase in
dissolved metal is also possible (Stoveland et al., 1979b; Rudd et al., 1988; Landner
and Reuther, 2004; Ziolko et al., in press). The mechanism of copper removal in
trickling filters is believed to be not dissimilar to that occurring in the activated sludge
process (Lester, 1987) however, although trickling filters are extensively used in the
UK, with over 800 STWs utilising them as a biological treatment stage in the South-
East region alone (Koh et al., 2008), they have not previously been studied in terms of
copper removal and therefore the exact mechanism of removal remains to be
established.
The biological treatment stage is increasingly followed by a range of different
types of tertiary treatment used to reduce biochemical oxygen demand, suspended
solids and ammonia concentration in effluents. Such treatment may possibly assist in
further reduction of copper concentrations in final effluents, however, these processes
were not specifically designed to remove copper, and their effectiveness is not well
documented (Kobler and Boller, 1997; Hamoda et al., 2004).
The aim of this study was to determine the concentrations of both total and
dissolved copper in the sewage at the different stages of sewage treatment at a number
of STWs in the North West of England in order to evaluate the effectiveness of
various unit treatment processes in removing copper.
2. Methodology
2.1. Description of STWs and sampling strategy
Eight STWs were sampled between November 2006 and November 2007.
Biological treatment at seven sites consisted of trickling filters (these locations are
5
identified as TF1 to TF7), and one site utilised an activated sludge plant, identified as
ASP. The sampling dates, site characteristics and their operational parameters, such as
proportion of returned liquors and recirculation over trickling filters, together with
details of any tertiary treatment processes utilised at each STW are presented in Table
1. All catchments were predominantly domestic with less than 11% of the total copper
load to the STWs from named traders, with the exception of TF3 where trade input
was estimated to be 20% of total load in crude sewage (Pers. Comm. United Utilities).
Samples of the influent and effluent from each unit treatment process were
taken every four hours from 10:00 on Monday through to 10:00 on Friday. Further
sampling points were added where samples could beneficially be taken, such as
combined flows of crude sewage with returned liquors or settled sewage with
recirculated flows (Figs. S1-S8 in supplementary material). Samples of sludges and
returned liquors, such as co-settled sludge from primary sedimentation, humus sludge
(from secondary sedimentation following biological treatment), decant liquors from
sludge treatment or backwashes from tertiary treatment processes were collected
daily, or as dictated by operation at each site.
Table 1 Site characteristics and operational parameters recorded during the
sampling periods.
6
2.2. Sample preparation
For the analysis of the concentrations of total copper in the influents and
effluents, 40 ml samples were acidified with 2 ml of 70% Optima-grade nitric acid
(Seastar Chemicals Inc., Pittsburgh, PA, USA) and then transferred to digestion
vessels for microwave assisted acid digestion undertaken on-site according to EPA
Method 3015, utilising a CEM Mars 5 digester with XP-1500 Plus vessels (CEM,
Matthews, NC, USA). For sludge and returned liquors samples containing suspended
solids concentrations greater than 1000 mg/l, prior to digestion (EPA Method 3051)
samples were adjusted to 0.1 g/l of solids by dilution with Ultrapure water (Millipore
UK, Watford, UK), and further acidified by addition of 1 ml of Optima-grade nitric
acid.
Analysis for dissolved copper involved filtration of the sample through a 0.45
μm cellulose nitrate membrane filter (VWR International, Darmstadt, Germany)
combined with a glass micro-fibre GF/D pre-filter (Whatman, Maidstone, UK). Prior
to filtration of the bulk sample, filters were washed with approximately 5 ml of
sample, which was discarded. The filtrate (40 ml) was then acidified with 1 ml of
Optima-grade nitric acid in order to solubilise copper prior to quantification. For
sludge and returned liquors samples with high solids content, samples were
centrifuged at 5000g for 30 minutes prior to filtration.
All glassware and digestion vessels used during the work were subject to a 2-
stage cleaning procedure using 3% solution of Decon90 (Decon Laboratories,
Brighton, UK) and 3% solution of nitric acid (Merck, Darmstadt, Germany) and
rinsed with Ultrapure water after each step.
7
2.3. Quantification of copper
Copper analyses were performed with a model 4110 ZL Zeeman-corrected
graphite furnace atomic absorption spectrometer (Perkin-Elmer, Beaconsfield, UK)
using a copper hollow cathode lamp (324.8 nm wavelength, 15 mA) and AS-72
autosampler with 2 ml polyethylene disposable sample cups. Calibration was
undertaken using 1, 2, 5, 10, 20 and 50 µg/l freshly prepared copper standards. Prior
to analysis, where necessary, samples were diluted so as to allow for the detection of
copper in the linear range of the calibration curve (Sterritt and Lester, 1980).
Five batches of unspiked samples together with samples spiked at low and
high standard additions were analysed for both total and dissolved copper to
characterise the method performance. Samples for both total (unfiltered and digested)
and dissolved copper (filtered, but not digested) spiked at low and high additions
exhibited no significant differences between batches with an overall recovery of
100% for total and 90% for dissolved copper. Recoveries for dissolved copper may be
explained by matrix effects in undigested samples (Belarra et al., 2000). All reported
results were corrected for these recoveries.
2.4. Data analysis
Statistical analyses were performed with JMP® 7 (SAS Institute, Cary, NC,
USA). Variance in the measured concentration of dissolved copper in samples from
the inlet and outlet of the unit treatment processes was analysed using the ANOVA
test. The post-hoc Tukey’s test was than used to determine where a statistically
significant difference existed between mean concentrations before and after a
treatment process.
8
2.5. Mass balance calculations
The mean daily flux of total copper through the unit treatment processes at
each site was calculated using concentrations and flows from data collected over the
sampling period. Changes in the flows resulting from the addition of returned liquors,
recirculation over trickling filters or due to desludging and backwashing were taken
into account. Data for fluxes through the works in many cases were calculated by
summation of measured values where flows combined, or subtraction where flow
streams split and these values are enclosed in boxes when presented in the mass
balance diagrams. Due to the uncertainty regarding the exact volumes of sludges and
returns and wide variability in their solid content, daily copper fluxes through the
works were not calculated.
3. Results
3.1. Spatial and temporal variation in copper concentrations in the crude sewage
The average weekly concentrations of total and dissolved copper determined
in the crude sewage at the eight sites are presented in Fig. 1. At the ASP it was not
possible to obtain crude sewage prior to the addition of returned liquors and therefore
the total copper concentration in the crude sewage was calculated using the mass
balance approach described above, with no attempt made to estimate dissolved copper
concentrations due to uncertainty regarding partitioning. There was some evidence of
spatial variation in both total (14.8 µg/l at TF1 to 107 µg/l at TF4) and dissolved (3.0
µg/l at TF2 to 7.7 µg/l at TF4) copper concentrations. Copper in the dissolved phase
accounted for between 7% (TF4) to 30% (TF3) of the total metal in the crude sewage,
which demonstrates that not only the total concentrations of copper entering the
STWs, but also the distribution between particulate and dissolved phases may vary
considerably.
9
Fig. 1. Total (■) and dissolved (□) copper concentrations (µg/l) in crude sewage.
It was also apparent that there was temporal variation in the concentrations of
both total and dissolved copper in the crude sewage entering each of the treatment
works. A typical example of the dynamic nature of the crude sewage, which also
demonstrates the link between diurnal flow patterns and copper concentrations in dry
weather, is shown in Fig. 2. The distinct diurnal pattern can be distinguished with the
lowest concentrations of copper consistently observed at the 06:00 sampling point.
Even during wet weather, which occurred most frequently at TF3, with FTFT
throughout the week averaging 92% of the design capacity (Table 1), a diurnal flow
pattern remained evident. The increase in copper concentrations with the increasing
flow rate during the day time implied the presence of a copper source with
concentrations above those present in runoff. As estimated inputs from trade were
mininal, the implication is that this source was of domestic origin.
10
Fig. 2. - Temporal variation in total (■) and dissolved (□) copper concentrations (µg/l)
showing a diurnal pattern with flow (l/s) to full treatment (―) at TF5.
3.2. The flux of copper through the sewage treatment works
At all sites, before crude sewage entered the primary sedimentation tanks it
was combined with the returned liquors streams. The concentrations of copper present
in some of the returned liquors (Table 2) were up to two orders of magnitude above
those observed in crude sewage (Fig. 1), indicating the potential impact that returns
may have on the load to primary treatment. Concentrations of total copper varied
considerably, not only between different return streams, but also at different sites for
the same return, however, in all cases copper was predominantly associated with the
solid phase.
Obtaining representative samples for flows of combined crude sewage plus
returned liquors was, however, difficult as returned liquor flows were frequently
unpredictable and access to suitable sampling points restricted. Fluxes to the primary
sedimentation tanks were therefore calculated using appropriate data on
concentrations and flows or volumes and compared to measured values (where
0.0
5.0
10.0
15.0
20.0
25.0
30.0
35.0
40.0
Mondy 1
0:00
Mondy 1
4:00
Mondy 1
8:00
Mondy 2
2:00
Tuesday
02:00
Tuesday
06:00
Tuesday
10:00
Tuesday
14:00
Tuesday
18:00
Tuesday
22:00
Wed
nesday
02:0
0
Wed
nesday
06:0
0
Wed
nesday
10:0
0
Wed
nesday
14:0
0
Wed
nesday
18:0
0
Wed
nesday
22:0
0
Thursd
ay 02
:00
Thursd
ay 06
:00
Thursd
ay 10
:00
Thursd
ay 14
:00
Thursd
ay 18
:00
Thursd
ay 22
:00
Friday
02:00
Friday
06:00
Friday
10:00
Cop
per
Con
cent
ratio
n, µ
g/l .
0
50
100
150
200
Flow
, l/s
11
available) for calculating and validating the mass balances for each site, which are
available in the supplementary information (Figs. S1-S8). These figures present the
flux of total copper in kg/day through the works, with values determined from
measured concentrations in open text, and calculated values, from combing two or
more flow streams, presented in text boxes. This approach was reliable for
quantifying the flux of total metal, however, this method is unsuitable for quantifying
the flux of copper present in the dissolved phase due to uncertainty related to
partitioning when two or more flows combine.
Although the volume of returned liquors did not account for a large proportion
of flow to the STWs, on average 8% (Table 1), their impact on the flux of copper to
the primary sedimentation tanks was significant. Copper in the returns accounted for
27% (the ASP) to 66% (TF3) of the load of total metal to the primary sedimentation
tanks with the major component typically being associated with biological solids,
either humus returns or surplus activated sludge. At a number of sites, where
representative samples of crude sewage plus returned liquors were obtained (e.g. TF3,
TF4 and TF7), fluxes to primary sedimentation tanks which were based on measured
concentrations at this same point, compared favourably with the calculated loads
(Figs. S2, S4, S7). However, at TF1 and TF6 (Figs. S1, S6), it was apparent that
sampling did not “catch” the total load present in the combined crude sewage and
returned liquors, indicating the importance of calculating the load of the total copper
by combining the flux in crude sewage and returned liquors.
3.3. The effectiveness of unit treatment processes at removing copper
The average removal of total copper based on fluxes through the STW
(including the impact of relevant returns) and dissolved copper (based on measured
12
concentrations) during conventional primary and biological treatment are summarised
in Table 3. The removal efficiencies for total copper have been calculated based on
values obtained from fluxes through the STW and therefore they include copper in
returned liquors and recirculation flows (where relevant), whereas values for the
dissolved metal fraction are determined from measured concentrations which as
explained, may not have captured internal return flows within each STW. However, it
is known that concentrations of dissolved copper in the returns were relatively low
(Table 2) and may not have any significant impact.
Table 2 Mean weekly total and dissolved copper concentrations (µg/l) in return
flows.
Site Humus Return Decant Liquor Backwash
Total Diss. Total Diss. Total Diss.
TF1 2144 0.5 - - - -
TF2 6676 1.8 111 0.8 50 2.9
TF3 1854 1.5 1274 0.4 63 4.1
TF4 4689 3.9 809 1.2 663 3.7
TF5 375 4.0 249 1.1 - -
TF6 380 2.8 1616 2.6 34.5 4.3
TF7 48 2.6 129 2.2 - -
ASP 837a 1.7
a - - - -
a Surplus activated sludge
The removal of total copper across the primary sedimentation tanks was
generally consistent at all sites, ranging from 49% (TF4) to 60% (TF2), with the
exception of TF6, where removal averaged just 38% over the sampling period. In
addition to the removal of copper associated with solids, removal of dissolved copper
was also observed at TF1, TF3, TF5 and TF7, with efficiencies reaching up to 49%
(TF7). However, at the other sites, there was no difference between crude and settled
13
sewage concentrations. The variability in the removal of dissolved copper was
possibly influenced by both variations in the proportion of returned liquors added to
the crude sewage prior to primary sedimentation (Table 1) as well as differences in
the nature of the returned liquors, such as redox potential or suspended solids
concentration, which may have influenced subsequent partitioning during primary
sedimentation.
During biological treatment, at the majority of the sites utilising trickling
filters removal of total copper was fairly consistent (Table 3), ranging from 44%
(TF7) to 55% (TF5), although, the potential for variability was highlighted by two
sites exhibiting more extreme removals of 22% (TF1) and 71% (TF3). The impact of
the biological trickling filters on the concentrations of dissolved copper
concentrations was not significant at most sites. However, at TF1 and TF5 significant
differences between dissolved copper in settled sewage and the effluent from the
biological treatment process indicate that substantial dissolution of copper occurred
(76% and 94% respectively), highlighting the unpredictable nature of this treatment
process in the removal of dissolved copper. It is possible that the rate of the
recirculation over the trickling filters, which varied considerably (Table 1), may have
changed the characteristics of the settled sewage by influencing redox potential
(dissolved oxygen concentration) and decreasing suspended solids, which could in
turn affect partitioning.
It was apparent that some of the tertiary treatment processes had the potential
to further reduce the total copper load discharged to the environment in final effluents,
although, their performance in this respect was variable. This was highlighted by the
variable performance of the BAFF processes installed at TF4 and TF6 sites (Table 4).
All tertiary processes designed to remove solids (RGF, COUF and BAFF) reduced the
14
load of total copper, probably as a result of decreasing suspended solids load,
although only the BAFF at TF4 removed dissolved copper, and at TF2, the COUF
was observed to increase concentrations of dissolved copper.
Table 3 Average removal efficiencies (%) of total and dissolved copper across
conventional unit treatment processes at each site. Negative values indicate an
increase in concentration.
Site Primary treatment Biological treatment
Total Dissolved Total Dissolved
TF1 53 13 22 -76 a
TF2 60 0 52 17
TF3 59 41 a 71 16
TF4 49 -1 50 -5
TF5 54 28 a 55 -94
a
TF6 38 -7 52 -7
TF7 53 49 44 -17
ASP 59 - 94 50 a
a Statistically significant (p=0.05)
Fig. 3 – The overall removal efficiencies from crude sewage to final affluent of total
(■) and dissolved (□) copper at the eight STWs. Negative removals indicate an
increase in concentration was observed.
-60
-40
-20
0
20
40
60
80
100
TF7 TF2 TF3 TF4 TF6 ASP
TricklingFilters
Trickling Filters withTertiary Treatment
TF1 TF5
Rem
ov
al e
ffic
ien
cy f
or
cop
per
(%)
15
Table 4 Average removal efficiencies (%) of total and dissolved copper across
tertiary treatment units. Negative values indicate an increase in concentration.
Site NTF BAFF RGF/COUF
Total Dissolved Total Dissolved Total Dissolved
TF2 -15 20 - - 45 -50 a
TF3 -15 -3 - - 32 -3
TF4 - - 61 52 a - -
TF6 - - 33 -2 - - a Statistically significant (p=0.05).
The overall performance of each complete STW, in terms of the removal
efficiency calculated between crude sewage and final effluent, are displayed in Figure
3. Although an attempt has been made to present the sites grouped by the treatment
processes present, there is no clear benefit displayed for the removal of copper when
tertiary processes are present. It is apparent, however, that the ASP was most effective
in removing total copper, but as crude sewage could not be sampled here, no value
could be calculated for removal of dissolved copper.
3.4. Concentrations in final effluent
The average concentrations of total and dissolved copper in the final effluents
from the STW ranged from 1.0 µg/l (dissolved) at the ASP to 15.2 µg/l (total) at TF4
(Fig. 4). As observed in crude sewage, there was a significant spatial variation in both
dissolved and total copper concentrations. However, the phase distribution in the final
effluents had markedly changed, in comparison to crude sewage, with a higher
proportion of copper being present in the dissolved phase and accounting for 26%
(TF4) up to 77% (TF5) of the total metal concentration. Additionally, the treatment
smoothed the temporal variation initially present in the crude sewages and there was
no diurnal pattern visible in the final effluent concentrations. In this set of data, the
16
quality of the effluent from the ASP was noticeably superior, even when compared
with the final effluents from trickling filter works incorporating tertiary treatment
processes.
4. Discussion
At most of the sites investigated, the concentrations of copper present in the
crude sewage were in the lower range of values recently reported (Karvelas et al.,
2003; Rule et al., 2006; Olivieira et al., 2007). Industrial inputs were not significant at
the sites studied, and diurnal variations indicated domestic users may be an important
source of copper to the STWs. Leaching from copper pipes is considered to be the
primary source of copper to STWs from domestic and commercial areas (Bouley and
Edwards, 2000; Sorme and Lagerkvist, 2002; Landner and Reuther, 2004). There is,
however, evidence that inputs from plumbing are decreasing, due to phosphate dosing
of potable water, which is known to significantly reduce the leaching of copper from
plumbing into the water supply (Edwards et al., 2002; Goh et al., 2008) and it is
known that phosphate dosing of potable supplies in the area studied was instigated
between 1997 and 1999 (United Utilities, 2001).
Fig. 4 - Total (■) and dissolved (□) copper concentrations (µg/l) in final effluents.
0
2
4
6
8
10
12
14
16
18
Cop
per
con
centr
atio
n,µ
g/
l
TF1 TF2 TF3 TF4 TF5 TF6 TF7 ASP
17
Regardless of the source, during this study copper in the influent was
predominantly found in the particulate form (70-93%) i.e. associated with solid
matter. This is in agreement with observation by other authors (Goldstone et al.,
1990a; Karvelas et al., 2003; Buzier et al., 2006), and therefore, copper was expected
to be efficiently removed during the primary sedimentation process. Moreover, the
addition of retuned liquors that typically contained less than 1% of copper in the
dissolved form further increased the ratio of total to dissolved metal. The removal
efficiencies of total copper observed in primary treatment during this study averaged
53% and were generally in the higher range of reported values (Olivier and Cosgrove,
1974; Kempton et al., 1987a; Goldstone et al., 1990a; Karvelas et al., 2003).
However, removal of dissolved copper during primary sedimentation was also
observed at three STWs. This phenomenon appears to be a consequence of the
addition of the returned liquors to the crude sewage. Retuned liquors typically contain
a large amount of suspended solids that may enhance partitioning of dissolved copper
onto the solid phase. Such behaviour has been reported previously (Goldstone et al.,
1990a; Buzier et al., 2006), although there is also evidence that the addition of
returned liquors may have the effect of increasing the concentration of dissolved
copper in the settled sewage (Brown et al., 1973). An increase in suspended solids
through the addition of returned liquors may also positively affect the removal of
copper associated with particles by increasing flocculation (Stoveland and Lester,
1980; Kempton et al., 1987b). However, variability in removal efficiencies
demonstrated that although the addition of returned liquors potentially enhanced
copper removal, site specific factors such as the nature / type of returned liquors may
have an impact on the fate of copper during primary treatment.
18
The removal efficiency of total copper during biological treatment varied from
22% to 71% on sites utilising trickling filters. Although during this study, the one site
utilising the activated sludge process performed considerably better (94%) than the
trickling filters, the removal values reported for this process also vary widely, from
25% (Roberts et al., 1977) through 43% (Karvelas et al., 2003) and 61% (Goldstone et
al., 1990a) up to 79% (Lester et al., 1979). This implies that although biological
treatment processes have the potential to significantly remove copper, the ultimate
performance is highly site specific and may depend on both physico-chemical
characteristics of the sewage (Sterritt et al., 1981) and operational parameters (Sterritt
and Lester, 1981). If dissolution of copper occurs across the biological process, this is
likely to reduce the effectiveness of any tertiary treatment process which reduces the
suspended solids in reducing the final concentration of copper discharged.
5. Conclusions
There was a wide temporal variability in copper concentrations in crude
sewage, however, they tend to follow the variation in diurnal flow which strongly
suggest they are indicative of copper originating from the water supply systems.
There is a wide variation in the removal efficiencies of copper between sites,
and this is especially evident for the dissolved metal fraction. Operational parameters
such as returned liquors and recirculation over trickling filters may all have an impact
on the removal of copper.
As the majority of the copper present following biological treatment was in the
dissolved phase, optimisation of existing conventional treatment processes (primary
and biological treatment) to maximise dissolved copper removal may have a
substantial impact on the overall removal efficiency of the STW and be a more cost
19
effective option than installing any tertiary process to specifically reduce the
discharge of copper.
Acknowledgements
The authors are grateful to United Utilities PLC for funding the work and for
permission to publish the results
References
Belarra MA, Crespo C, Resano M, Castillo JR. Direct determination of copper and
lead in sewage sludge by solid sampling-graphite furnace atomic absorption
spectrometry } study of the interference reduction in the gaseous phase working in
non-stop flow conditions. Spectrochim Acta B 2000; 55: 865-874.
Bouley N, Edwards M. Copper in urban water cycles. Crit Rev Environ Sci Technol
2000; 30: 297-326.
Brown HG, Hensley CP, McKinney GL, Robinson JL. Efficiency of heavy metals
removal in municipal sewage treatment plants. Environ Technol Lett 1973; 5: 103-
114.
Brown MJ, Lester JN. Metal removal in activated sludge: the role of bacteria
extracellular polymers. Water Res 1979; 3: 817-837.
Buzier R, Tusseau-Vuillemin MH, dit Meriadec CM, Rousselot O, Mouchel JM.
Trace metal speciation and fluxes within a major French wastewater treatment plant:
Impact of the successive treatments stages. Chemosphere 2006; 65: 2419-2426.
de Oliveira-Filho EC, Lopes RM, Paumgartten FJR. Comparative study on the
susceptibility of freshwater species to copper-based pesticides. Chemosphere 2004;
56: 369-374.
de Schamphelaere KAC, Vasconcelos FM, Tack FMG, Allen HE, Janssen CR. Effect
of dissolved organic matter source on acute copper toxicity to Daphnia magna.
Environ Toxicol Chem 2004; 23: 1248-1255.
Edwards M, Hidmi L, Gladwell D. Phosphate inhibition of soluble copper corrosion
by-product release. Corrosion Sci 2002; 44: 1057–1071.
European Commission. Directive 2000/60/EC of the European Parliament and of the
Council of 23 October 2000 establishing a framework for Community action in the
field of water policy. OJ L327; 22.12.2000. 2000,72 pp.
Goh KH, Lim TT, Chui PC. Evaluation of the effect of dosage, pH and contact time
on high-dose phosphate inhibition for copper corrosion control using response
surface methodology (RSM). Corrosion Sci 2008; 50: 918-927.
Hamoda MF, Al-Ghusain I, Al-Mutairi NZ. Sand filtration of wastewater for tertiary
treatment and water reuse. Desalination 2004; 164: 203-211.
20
Karvelas M, Katsoyiannis A, Samara C. Occurrence and fate of heavy metals in the
wastewater treatment process. Chemosphere 2003; 53: 1201-1210.
Kempton S, Sterritt RM, Lester, JN. Factors affecting the fate and behaviour of toxic
elements in the activated sludge process. Environ Pollut (Series A) 1983; 32: 51-78.
Kempton S, Sterritt R, Lester JN. Heavy metal removal in primary sedimentation. I.
The influence of metal solubility. Sci Total Environ 1987a 63: 231-246.
Kempton S, Sterritt R, Lester JN. Heavy metal removal in primary sedimentation. II.
The influence of metal speciation and particle size distribution. Sci Total Environ
1987b; 63: 247-258.
Kobler D, Boller M. Particle removal in different filtration systems for tertiary
wastewater treatment: A comparison. Water Sci Technol 1997; 36: 259-267.
Koh YKK, Chiu TY, Boobis A, Cartmell E, Pollard SA, Scrimshaw MD, Lester JN.
A sensitive and robust method for the determination of alkylphenol polyethoxylates
and their carboxylic acids and their transformation in a trickling filter wastewater
treatment plant. Chemosphere 2008; 73: 551–556.
Landner L, Reuther R. Metals in society and in the environment. A critical review of
current knowledge on fluxes, speciation, bioavailability and risk for adverse effects
of copper, chromium, nickel and zinc. Kluwer Academic Publishers, Netherlands,
2004, 406 pp.
Lawson PS, Sterritt RM, Lester JN. The speciation of metals in sewage and activated
sludge effluent. Water Air Soil Pollut 1984a; 21: 387-402.
Lawson PS, Sterritt RM, Lester JN. Adsorption and complexation mechanisms of
heavy metal uptake in activated sludge. J Chem Technol Biotechnol 1984b; 34B:
253-262.
Lawson PS, Sterritt RM, Lester JN. Factors affecting the removal of metals during
activated sludge wastewater treatment. I. The role of soluble ligands. Archives of
Environ Contam Toxicol 1984c; 13: 383-390.
Lawson PS, Sterritt RM, Lester JN. Factors affecting the removal of metals during
activated sludge wastewater treatment. II. The role of the mixed liquor biomass.
Archives of Environ Contam Toxicol 1984d; 13: 391-402.
Lester JN. Editor. Heavy Metals in Wastewater and Sludge Treatment Processes.
Volume II. Treatment and Disposal. CRC Press Inc., Boca Raton, USA, 1987, 155
pp.
Lester JN, Harrison RM, Perry R. The balance of heavy metals through a sewage
treatment works. I. Lead, cadmium and copper. Sci Total Environ 1979; 12: 13-23.
Liu Y, Lam MC, Fang HHP. Adsorption of heavy metals by EPS of activated sludge.
Water Sci Technol 2001; 43: 59-66.
Metcalf and Eddy Inc. Wastewater Engineering: Treatment and Reuse. 4th
Edition.
McGraw-Hill, 2004, pp. 361-411.
Oliveira AD, Bocio A, Beltramini Trevilato TM, Magosso Takayanagui, AM,
Domingo JL, Segura-Muñoz SI. Heavy metals in untreated/treated urban effluent
and sludge from a biological wastewater treatment plant. Environ Sci Pollut Res
2007; 14: 483-489.
Oliver BG, Cosgrove EG. The efficiency of heavy metal removal by a conventional
activated sludge treatment plant. Water Res 1974; 8: 869.
21
Roberts P, Hegi HR, Webber A, Kradhenbahl, HR. Metals in municipal waste water
and their elimination in sewage treatment. Prog Water Technol 1977; 8: 301-306.
Rudd T, Sterritt RM, Lester JN. Stability constants and complexation capacities of
complexes formed between heavy metals and extracellular polymers from activated
sludge. J Chem Tech Biotech 1983a; 33A: 374-380.
Rudd T, Sterritt RM, Lester JN. Mass balance of heavy metal uptake by encapsulated
cultures of Klebsiella aerogenes. Microb Ecol 1983b; 9: 261-272.
Rudd T, Lake DL, Mehrota I, Sterritt RM, Kirk PWW, Campbell JA, Lester JN.
Characterisation of metal forms in sewage sludge by chemical extraction and
progressive acidification. Sci Total Environ 1988; 74: 149-175.
Rule KL, Comber SDW, Ross D, Thornton A, Makropoulos CK, Rautiu R. Survey of
priority substances entering thirty English wastewater treatment works. Water
Environ J 2006; 20: 177-184.
Santarsiero A, Veschetti E, Donati G, Ottaviani M. Heavy Metal Distribution in
Wastewater from a Treatment Plant. Microchem J 1998; 59: 219-227.
Sorme L, Lagerkvist R. Sources of heavy metals in urban wastewater in Stockholm.
Sci Total Environ 2002; 298: 131-145.
Sterritt RM, Lester JN. Atomic absorption spectrophotometric analysis of the metal
content of waste water samples. Environ Technol Lett 1980; 1: 402-417.
Sterritt RM, Brown MJ, Lester JN. Metal removal by adsorption and precipitation in
the activated sludge process. Environ Pollut (Series A) 1981; 24: 313-323.
Sterritt RM, Lester JN. The influence of sludge age on heavy metal removal in the
activated sludge process. Water Res 1981; 15: 59-65.
Stoveland S, Astruc M, Lester JN, Perry R. The balance of heavy metals through
sewage treatment works. II. chromium, nickel and zinc. Sci. Total Environ.,
1979a, 12, 25-34.
Stoveland S, Lester JN, Perry R. The influence of nitrilotriacetic acid on heavy metal
transfer in the activated sludge process. I. at constant loading. Water Res 1979b;
13: 949-965.
Stoveland S, Lester JN. A study of the factors which influence metal removal in the
activated sludge process. Sci Total Environ 1980; 16: 37-54.
UKTAG (UK Technical Advisory Group on the Water Framework Directive).
Proposals for environmental quality standards for Annex VIII substances. SR1-
2007, (Revised June 2008), 2008, 77 pp.
United Utilities (United Utilities Water PLC). Drinking water compliance report
2001. UK, 2001, 23 pp.
Ziolko D, Martin OV, Scrimshaw MD, Lester JN. Evaluation of metal removal during
conventional wastewater treatment: the potential to achieve more stringent final
effluent standards. Crit Rev Environ Sci Technol; in press.
22
Table headers
Table 1 Site characteristics and operational parameters recorded during the
sampling periods.
Table 2 Total and dissolved copper concentrations (µg/l) in return flows.
Table 3 Average removal efficiencies (%) of total and dissolved copper across
conventional unit treatment processes at each site. Negative values indicate an
increase in concentration.
Table 4 Average removal efficiencies (%) of total and dissolved copper across
tertiary treatment units. Negative values indicate an increase in concentration.
Figure legends
Fig. 1. - Total (■) and dissolved (□) copper concentrations (µg/l) in crude sewage.
Data for the ASP is calculated from mass balances.
Fig. 2. - Temporal variation in total (■) and dissolved (□) copper concentrations (µg/l)
showing a diurnal pattern with flow (l/s) to full treatment (―) at TF5.
Fig. 3 – The overall removal efficiencies from crude sewage to final affluent of total
(■) and dissolved (□) copper at the eight STWs. Negative removals indicate an
increase in concentration was observed.
Fig. 4 - Total (■) and dissolved (□) copper concentrations (µg/l) in final effluents.
